Key result
A CNN-based QSAR model predicted cardiotoxicity (hERG blockade) with a correlation coefficient (R2) of 0.70, mean squared error of 0.62, and mean absolute error of 0.65 on the test dataset.
Why the study?
Small chemicals blocking potassium ion channels prolong the QT interval and cause cardiotoxicity, making quantitative prediction of hERG blockers essential for drug development.
A novel CNN-based QSAR model demonstrates strong predictive performance for hERG-related cardiotoxicity, providing a useful computational tool for early drug development.
Supports AI-based hERG screening in drug development; leaves open clinical translation without prospective validation.
Small chemicals that block a potassium ion channel result in a prolonged QT interval, which can have serious cardiotoxic effects and is a major factor in drug development failures. To develop the drug successfully, quantitative prediction of human-ether-a-go-go-related (hERG) blockers is essential for designing drug candidates without the risk of cardiotoxicity. We built a convolutional neural network (CNN)-based quantitative structure–activity relationships (QSAR) model to predict cardiotoxicity. The statistical parameters of mean squared error (MSE) were 0.001, the mean absolute error (MAE) was 0.016, and the correlation coefficient (Q 2 ) was 0.99 for the training dataset. The MSE was 0.62, the MAE was 0.65, and the predicted correlation coefficient (R 2 ) was 0.70 for the test dataset. Further, we explored principal component (PC) analysis, t-SNE, scaffold analysis, active cliff, fingerprint analysis and chemical analyses to identify molecular similarity. We discovered that adding an acidic oxygen/aliphatic oxygen (hydroxyl group) reduces hERG inhibition and increases lipophilicity. The fragments are furan, sulfonamide, methanesulfonamide, p-chlorophenyl, p-fluorophenyl, and ethyl(heptyl)amino groups increased the hERG risk. Finally, we conclude that the QSAR model in combination with the convolutional neural network (CNN) offers a potentially novel approach for quantitatively predicting the cardiotoxicity of drug candidates.
No takes yet. Share an insight, caveat, or question.
Kumari et al. (2025) studied Cardiotoxicity. CNN-based QSAR model was evaluated on Prediction of cardiotoxicity (hERG blockers) (R2 0.70). A CNN-based QSAR model predicted cardiotoxicity (hERG blockade) with a correlation coefficient (R2) of 0.70, mean squared error of 0.62, and mean absolute error of 0.65 on the test dataset.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: